TRUE STEREO IR'S

Where can I find "true stereo IR's. I know about Noisevault- are there any other sites like that.
I'm using Pristine Space.
Thanks.

This is a nice collection for a reasonable price. Several of the IRs have true stereo files.

http://www.studiocat.com/index_files/Page1240.htm

ORIGINAL: John

What would be the advantage to a stereo IR?

With a mono in-stereo out reverb, the reverb return is the same right and left, no matter where you pan the source. So-if you pan hard left, you still hear reverb ar the right side.
On a stereo in/stereo out reverb you get a true "reflection" of the reverb compared to where you are in the R-L Field.
Panning the mono in/stereo out reverb is not the same- that doesn't really give you that sense of space.

ORIGINAL: Hansenhaus

This is a nice collection for a reasonable price. Several of the IRs have true stereo files.

http://www.studiocat.com/index_files/Page1240.htm

Hey Eric,

Just wanted you to know your CDs have been doing rotation in My car for the last couple weeks:)
Great stuff dude.

Tim

Sorry, wrong thread...

Greetings all,

The problem being discussed here is one of the reasons I prefer to do acoustical modelling for serious works.

I move the source around on a semicircle, calculating IR's for each position (left and right) using either the pressure or the gradient of the pressure (omni for binaural and directional for Blumlein, respectively). There is no panning when I do this --- it would cause the problem mentioned in this thread. Instead the amplitudes change naturally. The IR's are used one pair for each part, and this establishes BOTH direct wave (dry) and ambients (reflections and "reverb") rather than faking it like true stereo. True stereo is a compromise between what I do and what (say) Perfect Space does with panning. Some may argue that the ambience doesn't change markedly anyway, so faking it is OK. This is not quite true. It depends on the room, etc. In my typical setups, the positions to right and left are significantly closer to the walls, so early reflections are modified quite a bit. In binaural, it's particularly important to get things correct or they don't sound right.

Regards,
Dave Clark

P.S. On edit: I should have said, in case it's not obvious, that the load with what I do is even larger. I can easily be doing 7 stereo convolutions --- not exactly real-time yet with a lot of soft synths going, but getting closer with these multiprocessor CPU's.

Ok so far I am not sure that I understand the need for stereo IRs in the way I use PS. I use it as the last reverb in a full mix and with just a touch of it to give an ambiance to the whole mix. For single track or gropes I use the Pantheon or some other reverb. I also use FX 3 if I wish to have a more positional reverb. There is also surround that can be used to get positioning for a stereo file. I can't see the need for a stereo IR in processing say a guitar amp or other mono type effects. Mention was made to panning hard left a mono source and thus getting a hard left output. I don't see that as a problem or a deficit in how the convoluter works. Basically I am not sure why it would be desirable to use two separate IRs in a convoluter FX in order to recreate a spatial effect. Also I am really at a loss to understand how PS could use a stereo IR any better then it can with a mono IR. Its sort of the nature of the FX to use the mono IR to recreate the acoustical space it was made from. Wouldn't a stereo version add nothing more then a lot more processing?

I am still open to learning what a stereo IR can offer me that I can't already do with the tools I have at hand.

I don't think Perfect Space is a candidate for using a True Stereo IR. Maybe that's the basis of why you don't see an advantage to it.

I was aware of that. One reason I wonder about it. If it is so needed why not in PS?

Mike I do appreciate what you are saying as well as your tact. What I am not clear about is how the added data of a stereo IR is going to make a better rendition of an acoustical space then a normal IR? After all isn't that what a convoluter is for in the first place? Wouldn't it be dependent on algorithms of the convoluter to use the IR to create that acoustical space? Are they insufficient to do so? Is there some table or testing one could do short of buying an expensive convoluter to hear or see a difference? I guess what I am getting at here is as often the case, why is this needed and who benefits from it? Being around awhile I have seen lots of things that purport to be"better" but aren't. The idea of an IR by its nature seems to rebuke the notion. Unless all non stereo convoluters should be used only on mono sources and output only mono. We know that this is not so.

Don't the majority of reverb pluggs out there simply sum the inputs to mono , process them and build a new stereo image to output ??? I notice that they put the hundreds echos basically panned every freaking where possible so as to sound big ; but enough of this is going to turn things to mush pronto!!

I looked at the "True stereo " thing awhile back , And If you do the schematic and follow it , it's just like a cross delay in that when you pann things to one side , the crossed part of the path still allows the other to recieve some signal .

It's sounds like mike has a way of getting more control of where the echo's end up . Are you setting up many different mono reverbs ??


At the moment , I basically have a template with two FX busses set up for reverb , panned left and right ; then I use two sends per strip ( and you could pan those send signals too , the configuration is there !) to try and gain some control in where the echos from one instrument VS. another end up . It is more work , but keeps the ambience from being one big soup , I also use ducking sometimes to keep it under control .

Mike , Could you break you methodoligy down a bit more ( for someone like my pedantic self!!




Cheers)

Don't the majority of reverb pluggs out there simply sum the inputs to mono , process them and build a new stereo image to output

I should hope not. First and foremost that would be the single worst way to deal with real stereo. Secondly that would mean that all the positional data that is embedded with the tracks and sent to the reverb would be erased. Not to mention phase and other related data. What an awful way to have a darn reverb. Is it worth it?

Hi John,

I'm not sure if I understand what you are asking or not, but I'll try to provide something here for thought.

A single IR represents the response to ONE location of a source and ONE location of a receiver in a room. A stereo IR (not "True Stereo" but a stereo pair in Perfect Space) COULD represent one location and two ears. It could also represent one location and and X-Y pair of coincident directional microphones.

Your use of Perfect Space, as I understood your post, was to mix several locations and (say) an X-Y pair of coincident directional microphones, the mixing being done by mere panning that you do in the mixer. This mixing is not physically correct. It is incorrect because only one IR pair is being used (at most) in Perfect Space, but you are using it to represent many different locations in a room (the virtual room or studio that the panning creates).

Example: For such an X-Y pair, consider the guitarist at 45 degrees to the left. No direct wave will be recorded in the right-channel mic. However, everything else (all other waves) will be just as loud as in the left side. When you pan, this isn't true, so it doesn't sound right after Perfect Space. The ambiance should be just as loud in both channels.

Continuing the example, consider the bassist straight ahead. A reduced direct wave will be recorded in both channels, roughly at the same level. Everything else should also be roughly the same (depending on where the walls are, etc.). After Perfect Space, this will sound about right if you pan to Center.

The result of using only one IR in Perfect Space is that the bassist is correct, but the guitarist is wrong. If you're using a very slight reverb, or if you're listening to this in a room that has its own reverb, you may not notice this. It's still wrong.

Hope that helps!

Best regards,
Dave Clark

P.S. On edit: Jim Roseberry posted at almost the same time below --- same idea.

I am still open to learning what a stereo IR can offer me that I can't already do with the tools I have at hand.

Hi John,

Most algorithmic reverbs are true stereo... but they often don't provide the density of a world-class hardware unit. Convolution reverb can do this (at the expense of not representing modulated tails - for which Lexicon is famous).

If you're standing in a room listening to a guitar player... and the guitarist (and his rig) are to your far left, you'll hear the room ambience in both ears (not just your left). :)
Recreating this scenario with a guitar track and convolution plugin: With a stereo (as opposed to true stereo) IR, you'd only hear the ambience in the left channel. Nothing like the real thing...

.

I wish I could say I knew more about this but I don't. I do know some things about stereo that come from long ago. That is why I asked the question. I vaguely understand how an impulse can be used to reproduce the acoustic space of a room in a convolution plugin. I don't understand the math or how it is done. But I do know that it works. I just don't get how a stereo impulse is going to add anything to the process.

If you know some of my past posts on stereo you will know I do have a background on binaural recording. A real stereo recording should not have a reverb involved anyway. At least not post recording.